[Intel-gfx] [PATCH 04/23] drm: Add drm structures for palette color property

[Ville Syrjälä]

On Tue, Sep 22, 2015 at 02:53:54PM +0100, Emil Velikov wrote:
> On 22 September 2015 at 14:08, Daniel Vetter <daniel at ffwll.ch> wrote:
> > On Wed, Sep 16, 2015 at 11:07:01PM +0530, Shashank Sharma wrote:
> >> From: Kausal Malladi <kausalmalladi at gmail.com>
> >>
> >> This patch adds new structures in DRM layer for Palette color
> >> correction.These structures will be used by user space agents
> >> to configure appropriate number of samples and Palette LUT for
> >> a platform.
> >>
> >> Signed-off-by: Shashank Sharma <shashank.sharma at intel.com>
> >> Signed-off-by: Kausal Malladi <kausalmalladi at gmail.com>
> >> ---
> >>  include/uapi/drm/drm.h | 27 +++++++++++++++++++++++++++
> >>  1 file changed, 27 insertions(+)
> >>
> >> diff --git a/include/uapi/drm/drm.h b/include/uapi/drm/drm.h
> >> index e3c642f..f72b916 100644
> >> --- a/include/uapi/drm/drm.h
> >> +++ b/include/uapi/drm/drm.h
> [snip]
> >> +struct drm_palette {
> [snip]
> > ... extending the ioctl struct at the end ...
> Is this really going to work, considering that we currently have a
> zero sized drm_r32g32b32 array at the end ?

Well in this particular case it would be ugly. Sure, it can be done,
but we couldn't actually define the struct for it using C. Would be
a neat feature though. Someone should propose it for the next C
standard :)

In any case, for blobs I think we shouldn't extend them like ioctls.
In this case the blob is just an array of something, so the blob size
can of course vary depending how many elements are required for the
particual platform.

> Iirc someone mentioned that using a pointer to the data (over an
> array) has drawbacks, although for the sake of me I cannot think of
> any. Can anyone care to enlighten me, please ?

I guess an extensible array at the end of the struct is simply a nice
fit sometimes. Also makes it more obvious how much junk gets copied over
I suppose.

-- 
Ville Syrjälä
Intel OTC